Why so many Core i7-2600 flavours?
Posted on 9th Aug 2011 at 11:35 by Clive Webster with 26 comments
There’s an interesting article over at Ars Technica, titled What processor should I buy: Intel’s crazy pricing makes my head hurt. That might seem a silly question at first: as the author points out, surely you just buy the most expensive CPU in the LGA1155 range. However, Peter Bright is no fool; looking closer at the specs and his requirements, the author struggles to make sense of Intel’s strategy with new features, performance and compatibility.
The problem is due to Bright’s desire to make a future-proof, fast PC that can run Visual Studio and Battlefield 3 easily. A Core i7-2600 is a no-brainer, but there are three flavours, with the S model even running at slower stock speeds to save 30W of power (it Turbo Boosts to the same 3.8GHz as the other i7-2600 CPUs, however).
Then there’s the toss-up between the i7-2600 and the i7-2600K – the former has some interesting virtualisation and security features that Bright wants, but the latter has a better GPU and the ability to overclock. So which one is better? They both seem compromised and yet there’s a £10 ($23) price difference. The point is really, why has Intel disabled the useful VT-d and the potentially useful TXT logic from the i7-2600K?
Sure, TXT could be seen as a way to introduce hardware-based DRM to a home PC, but as Bright points out, it could also be very useful in preventing rootkits from slaving your PC to their nefarious desire (my melodramatic wording, not his).
Bright finds a solution to his quandary in the Xeon world, where there is a CPU that fits his needs, but then he’s stymied by the lack of Smart Response on official Xeon chipsets. So he’ll have to opt for the not officially supported combination of a Z68 motherboard with a Xeon processor. This should work fine, but for a PC you absolutely rely on for work (I assume) this isn’t a comfortable arrangement.
So why does Intel feel the need to disable potentially useful features from its supposedly top-end CPU when this will slow down uptake? And is the lack of Smart Response technology in any Xeon chipset a tacit admission that it’s not 100 per cent reliable? Conspiracy theories below please!
The problem is due to Bright’s desire to make a future-proof, fast PC that can run Visual Studio and Battlefield 3 easily. A Core i7-2600 is a no-brainer, but there are three flavours, with the S model even running at slower stock speeds to save 30W of power (it Turbo Boosts to the same 3.8GHz as the other i7-2600 CPUs, however).
Then there’s the toss-up between the i7-2600 and the i7-2600K – the former has some interesting virtualisation and security features that Bright wants, but the latter has a better GPU and the ability to overclock. So which one is better? They both seem compromised and yet there’s a £10 ($23) price difference. The point is really, why has Intel disabled the useful VT-d and the potentially useful TXT logic from the i7-2600K?
Sure, TXT could be seen as a way to introduce hardware-based DRM to a home PC, but as Bright points out, it could also be very useful in preventing rootkits from slaving your PC to their nefarious desire (my melodramatic wording, not his).
Bright finds a solution to his quandary in the Xeon world, where there is a CPU that fits his needs, but then he’s stymied by the lack of Smart Response on official Xeon chipsets. So he’ll have to opt for the not officially supported combination of a Z68 motherboard with a Xeon processor. This should work fine, but for a PC you absolutely rely on for work (I assume) this isn’t a comfortable arrangement.
So why does Intel feel the need to disable potentially useful features from its supposedly top-end CPU when this will slow down uptake? And is the lack of Smart Response technology in any Xeon chipset a tacit admission that it’s not 100 per cent reliable? Conspiracy theories below please!
26 Comments
Discuss in the forums Replythe K - overclocking, better igp, lacks some business-orientated bits [vPro, VT-d]
the vanilla
the S - lower TDP, lower speed [same turbo speed]
the T - even lower TDP, even lower speed [and lower Turbo], lower igp frequency
EDIT: mis-read, you're on about the 2600, the above still applies, except there is no T version of the 2600
Thats how intel does it.
competition Bulldozer can't get here soon enough
Intel realise a 2500k without IGP and use some of that production saving to cut prices by £10?
If your a gamer, overclock with a K
Thats about it really.
I've thought it was a bit obvious myself to be honest, with the letters representing different features. They're essentially different ranges, just a bit jumbled. From Wiki:
They should have put the letters before the name, but as I said, view them as separate ranges and it isn't a problem. So I completely disagree with the article on that front.
They should have left the virtualisation features on the K models, however.
heck even the g6950 has virtualization. I use it all the time too
Intel didn't design Sandy Bridge to be the jack-of-all-trades, this was obviously some marketing decision intended to increase volumes by having "specialized" processors. Either way, I've never been a fan of using a single computer for both work and game.There's just some comfort in having another computer with all your confidential files/designs/etc that your friends and family know not to touch.
They would save very little if any money. Chances are if the Chip is good enough to be highly overclockable (remember Intel are binning all their Sandybridges) then the IGP also works solidly. It would be pointless to disable bits of it if they're good chips, it wouldn't gain them anything or save them anything. At least this way, anyone who is just out for CPU performance (they do exist) also gets the gain of having all 12 EU's available, even if it is only for Video Encoding or the like. Makes more sense now that Z68 is out and we can see the bigger picture.
I see your point. Technically Intel's leading chip atm is the 990X which supports VT-d but not TXT. As has been said they're catering to more than just the consumer market and you've also got to put it in perspective as a relative scale. Of the S chips, this is the 2600 of the S Series, it supports such and such. The S for example, also has embedded options available and so really can't be put in the same box as other normal chips. All that considered it shares alot with the normal 2600. By the same logic you could say that the i7-2620M should be faster and have the same feature set as the desktop 2600's. But again it's relative to other M chips.
Anyway, on the feature front, what Intel are really saying is, If you want VT-d, TXT and Overclocking then you're going to have to pay "Extreme" prices for it when they're released, and they can do that because they have such massive market dominance in the high performance sector.
And can someone explain, why all of the notebook-parts have the HD3000 in that regard, while the desktop-parts do not?
It's all a total mess, and that doesn't count for intel alone.
Still for home user that is not a big deal. If you want to setup virtual machine (e.g. for XP) then there is always Virtual Box.
It's a masssive mistake for Intel not to support it on the K CPUs. Rather than dual-booting XP/7 as I currently do, with VT-d support I could simply install 7 & then install XP as a virtual PC with it having access to my GPU for gaming (and gaining from not having to accomodate XP with the correct BIOS settings (lack of ACHI support and NCQ on HDDs being a good example)). Surely disabling such an important feature on a CPU that is most likely going to be used in gaming rigs is a major fail...
The 2500k was never meant to have the disabled features so they turned them off.